In the comments of the last post, I was challenged to think a bit more carefully about categories, and more generally about my thinking style. This is not a response to those comments, but something motivated by me thinking about the topic.

I’m a messy thinker. I like to see masses of data, lots of overloaded axes, partial information, overlapping categories and contradiction. I like to see patterns arise out of the mess, and am less comfortable with reorganizing the mess into some form of order (though that is sometimes necessary).

Borges writes on categories (in El idioma analítico de John Wilkins, 1942)

“These ambiguities, redundancies, and deficiencies recall those attributed by Dr. Franz Kuhn to a certain Chinese encyclopaedia called the Heavenly Emporium of Benevolent Knowledge. In its distant pages it is written that animals are divided into (a) those that belong to the emperor; (b) embalmed ones; (c) those that are trained; (d) suckling pigs; (e) mermaids; (f) fabulous ones; (g) stray dogs; (h) those that are included in this classification; (i) those that tremble as if they were mad; (j) innumerable ones; (k) those drawn with a very fine camel’s-hair brush; (l) etcetera; (m) those that have just broken the flower vase; (n) those that at a distance resemble flies.”

This passage has been picked up by other writers, notably Foucault. Some who took Borges seriously on his identification of a source (a bad move, Borges is often fictive about scholarship) made rather racist points about Chinese conceptual systems. Others simply laughed at how terrible such a classification would be.

I think it sounds cool.

I think part of the gaffawing is because a lot of writers on this passage miss the implication of (h). It is not the case that each animal can appear in only one category. So this is a categorization based on a set of features that (though intended as humorous) are presumably all useful in some way.

Here are two categorizations:

I divide animals into these categories: vertebrates and invertebrates. Of the vertebrates I sub-divide them into mammals, birds, fish, reptiles and amphibians. The mammals I further divide into monotremes, marsupials and placental mammals.[1]


I divide animals into these categories: those I own, those that are dangerous, those I have in a book, those that can be eaten, those my nieces would like on a lunchbox, and those that I’d like to paint.

Which is better? The latter, it seems to me. Give me Borges’s encyclopaedia any day.

How about you?

PS: There are serious implications behind this. I do make these kinds of decisions regularly in my day job. I’ve been trained to think that hierarchies of categories are better, but I’ve time and time found that for real data and real human needs, ad-hoc mixed categories (or ‘tags’, which are a version of the same thing) are superior.

[1] Apologies for the biological naivety to anyone who would have expected better things given my background. But the naivety makes a related point: the first style of classification seems systematic, but is actually just as arbitrary.


Filed under Uncategorized

15 responses to “Categories

  1. TWF

    I’d definitely opt for the “fuzzy” categories over the scientific ones too. They’re far more practical outside of the world of rigorous science. 🙂

    Being the egocentric beasts that we are, I think the latter method is more natural to us too, because we are then defining things in terms of our perspectives as opposed to impersonal characteristics.

  2. Pingback: Perception – categorization | The Heretical Philosopher

  3. Which is better?

    Better for what?

    I thought your two different examples illustrated very well two different meanings for “categorize”. So I posted about that at my blog.

  4. Ian

    Thanks Neil

    Beter for what?

    In the post, I meant a better match for how I think about things.

    But in a very real sense, better for anything. Since the second kind of categorization is strictly more powerful than the first (in the sense that it can easily represent any first-type categorization, the reverse being untrue).

  5. “Categories” are tools. So it depends on your purpose as to what sort of categories are useful: fuzzy or razor-sharp.

    There is a fun hold on categories in science: we try to draw them as clearly sharp as we can so that we can test and then prove our mistakes. The problem exists for those who can’t do that.
    They are those who prefer fuzzy or prefer unmovable sharp borders. Scientists and laypeople alike error on both sides of these mistaken uses.

    And folks jump back and forth too. At one time they will demand more fuzzy and at other times, more defined. It depends what they are attempting to do.

  6. For giggles, and the opportunity to click follow this time, I looked up some etymology stuff:

    Taxis = arrangement
    Nomia = method (management)

    Cate (kata) = against
    agora = public assembly

  7. Ian

    @sabio – thanks.

    Or perhaps nomia = law, taxonomist = arrangement lawyer 🙂 Interesting that the root of category is a word for argument.

  8. Two category schemes I have argued against for years are the Myers-Briggs Personality classification and the DSM (used in psychiatry and medicine). People are tricked by categories often.

  9. Ian

    @sabio – MBTI I think is pretty obviously suspicious, but I’ve only ever heard DSM mentioned, so I can’t anticipate what you think is wrong there. Can you say more (or point to something you’ve written previously on it)?

  10. Ian:
    Here is one link:
    I think you know that blog.
    But there are many such criticisms out there. Maybe the DSM is not used in Europe. Or you didn’t know which DSM I was referring to (though I said, psychiatry and medicine).

  11. Ian

    Yes, that was the DSM I thought you meant. I haven’t ever read one, and I haven’t heard it talked about in generally skeptical terms. Not saying that to defend it, just pleading ignorance!. Thanks for the link – I’ll look now.

  12. I consulted the Liddell-Scott lexicon. A kategor was an accuser, a kategoria was an accusation, but in logic, a kategoria somehow began to mean “predicate”. ‘Twould be interesting to see a history of the word.

  13. Scott de B.

    The advantage of hierarchical categorization is that it tends to be easier to apply with consistency.

    If you have 20 people to categorize 10,000 things, and you want to make sure that they are all categorized in the same way, and minimizing errors both of commission and omission, I think a hierarchical categorization is the way to go.

  14. Ian

    Scott I don’t get why a hierarchical classification has this property, but an exclusive one doesn’t. Or do you just mean a classification that is agreed upon before the thing needing to be classified, as your second paragraph suggests?

  15. Ian

    @Dan, thanks for that – yes it would be interesting to see how it got its current nuance.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s